I'm Omar Williams — 14+ years building semantic models, DAX, and Fabric governance for FP&A, treasury and credit risk teams. I run Milestone BI, a practice built on one idea: numbers that reconcile the first time.
I'm Omar — I've spent the last 14+ years inside regulated institutions that cannot afford a wrong number: banks, asset managers, and insurers where a broken variance report isn't an inconvenience, it's a control failure. My focus sits at the intersection of Microsoft Fabric and financial services domain fluency — semantic modelling, DAX, and governed lakehouse architecture built specifically for FP&A, treasury and credit risk reporting.
Milestone BI is the practice I run, but the work is mine — every model, every measure, every governance decision is something I've built and stood behind personally. Consulting engagements double as design partnerships: real problems from real reporting teams inform the tools I build next, including an AI-assisted financial commentary engine I'm currently developing to automate variance narratives at month-end close.
Packaged engagements, not open-ended retainers — scoped so you know what you're getting before we start.
A focused engagement to rebuild your management reporting on a governed semantic model — accurate, fast, and reconciled to source.
Lakehouse and semantic-layer governance built for institutions that get audited — lineage, access control, and version discipline from day one.
Ongoing embedded BI leadership — model maintenance, DAX optimisation, and roadmap input, billed monthly.
Selected builds — from full star-schema Power BI solutions to prototype product work.
A portfolio-grade Power BI solution on a synthetic online retail dataset — full star schema, 30+ DAX measures, and a five-page report design covering sales, margin, and operational performance.
A cash flow dashboard for small businesses (1–10 staff) — runway forecasting, automated invoice chasing, and real-time P&L visibility in a single view.
Replaced a spreadsheet-based liquidity pack with a governed semantic model, cutting month-end reconciliation time significantly.
Built a credit risk exposure dashboard with full data lineage documentation to satisfy internal audit requirements.
Designed a DAX-driven variance engine that flags and explains material budget deviations before the close review.
Three phases, scoped up front — no open-ended discovery loops.
Audit current reporting, data sources, and where the numbers stop reconciling.
Star schema, DAX measures, and report pages built against agreed requirements.
Documentation, access control, and a team that can maintain it without me.